AUTHORS
Hui Jin, MD, PhD, Zihan Liang2, Zijuan Wu3, Yi Miao, MD4, Wenyu Shi5, Jianyong Li, MD6
AFFILIATIONS
1 The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
2 The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
3 The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
4 The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
5 Affiliated Hospital of Nantong University, Nanjing, China
6 The First Affiliated Hospital of Nanjing Medical University, Nanjing, China
ABSTRACT
Introduction
Peripheral T-cell lymphoma (PTCL) represents a heterogeneous group of mature T-cell malignancies with unfavorable clinical outcomes and limited therapeutic options. Reliable blood-based biomarkers for diagnosis and risk stratification remain scarce, and the mechanisms underpinning systemic immune dysregulation in PTCL are incompletely elucidated. We therefore performed integrative peripheral blood multi-omics profiling to construct a comprehensive molecular atlas of PTCL, develop diagnostic and prognostic stratification models, identify key molecular regulators, and characterize their biological functions and therapeutic potential.
Methods
We performed single-cell RNA sequencing (scRNA-seq) on peripheral blood mononuclear cells (PBMCs) from 72 patients with PTCL and 5 healthy donors, with paired single-cell T-cell receptor sequencing (scTCR-seq) conducted for 60 patients. In parallel, we generated data-independent acquisition (DIA)-based proteomic profiles and targeted metabolomic profiles from plasma samples of 79 PTCL patients and 49 healthy donors. To delineate the cellular origin of candidate biomarkers, we further performed scRNA-seq on tumor tissues from 16 patients with angioimmunoblastic T-cell lymphoma (AITL) and 3 normal lymphoid tissues. We constructed machine learning algorithms integrating plasma proteomic and metabolomic signatures for PTCL diagnosis and prognostic stratification. The biological function, cellular origin, spatial distribution, and therapeutic potential of candidate molecules were validated using public datasets (GSE160119 and GSE58445), flow cytometry, multiplex immunofluorescence analysis of tissue microarrays (TMAs) from 150 PTCL patients, and an EL4 syngeneic lymphoma mouse model.
Results
Integrative multi-omics analysis yielded a comprehensive peripheral blood atlas of PTCL for the forst time, encompassing single-cell transcriptomic, T-cell receptor repertoire, plasma proteomic, and metabolomic landscapes. Machine learning models for PTCL diagnosis and prognostic prediction were developed based on plasma proteomic and metabolomic signatures.
Notably, cross-platform integrative analyses consistently indicated that VSIG4 may participate in PTCL tumorigenesis and disease progression. Survival analysis in the independent GSE58445 cohort revealed that patients with high VSIG4 expression exhibited significantly inferior overall survival relative to those with low VSIG4 expression. Single-cell analysis of AITL tissues further indicated that circulating soluble VSIG4 predominantly originates from tumor-associated macrophages, instead of peripheral immune cells.
Integrated analyses combining public transcriptomic datasets, single-cell RNA sequencing (scRNA-seq) with paired BCR/TCR profiling, and peripheral blood flow cytometry consistently revealed that VSIG4+ monocytes/macrophages were associated with an exhausted CD8+ T-cell phenotype. Multiplex immunofluorescence-based spatial analysis further showed that CD8+ PD-1+ T cells preferentially reside in close proximity to CD68+ VSIG4+ tumor-associated macrophages relative to CD8+ PD-1- T cells, suggesting that VSIG4+ macrophages contribute to the establishment of an immunosuppressive niche that drives CD8? T-cell exhaustion.
In therapeutic validation, anti-VSIG4 blockade significantly enhanced the antitumor efficacy of PD-1 inhibition in the EL4 lymphoma model. Combined treatment with anti-VSIG4 and anti-PD-1 achieved superior tumor growth inhibition compared with monotherapy, supporting VSIG4 as a promising target for combination immunotherapy in PTCL.
Conclusions
Our study establishes the first comprehensive peripheral blood multi-omics atlas of PTCL and develops machine learning models based on plasma proteomic and metabolomic signatures for diagnosis and prognostic stratification. Integrative multi-omics analyses identify VSIG4 may act as a macrophage-derived immunoregulatory molecule driving PTCL progression. Mechanistically, VSIG4+ macrophages likely foster the formation of an immunosuppressive microenvironment by promoting CD8+ T-cell exhaustion, and therapeutic blockade of VSIG4 potentiates the antitumor efficacy of PD-1 blockade in vivo. These findings imply that VSIG4 holds promise both as a circulating biomarker and a prospective immunotherapeutic target for patients with PTCL.
DISCLOSURES
Hui Jin, MD, PhD: No relevant conflicts of interest to declare.